The FE405B AD plot I recently posted used a simple (non 50% duty cycle) divide by 3 using a 74ls74 and 74ls02 (if I remember correctly). I used the HP board from the pair used in the 5065A or 5061A/B. The existing 74ls74 was piggy backed with the 74ls02 and connections made with tiny wire. The non-symmetrical 5Mhz output of the divider went into the existing 5Mhz filter amp resulting in a nice clean Sine wave. As the plot shows the AD does not suffer using this simple scheme. Cheers, Corby